Emerging Distributed/Parallel Computing Systems

1. Introduction

Computing continues to evolve rapidly, driven by data-intensive applications, AI workloads, and increasingly heterogeneous, interconnected infrastructures. Distributed and parallel computing systems are foundational to this evolution, enabling scalable performance, resilience, and efficient resource utilization across cloud, edge, and cyber–physical deployments. This Special Issue, titled “Emerging Distributed/Parallel Computing Systems”, will highlight cutting-edge advances, emerging trends, and key challenges spanning architectures, algorithms, systems, and applications in distributed and parallel computing.

The Special Issue is now closed (deadline: 15 November 2025) and includes eight peer-reviewed papers—five research articles and three review papers—covering surveys of timely subfields, as well as research contributions to security, privacy, and learning-enabled data processing.

2. Overview of Published Contributions

A helpful way to understand the collection is through dividing its papers into four themes.

Theme A—Foundations and surveys: trends, programmability, and encrypted visibility

“State of the Art in Parallel and Distributed Systems: Emerging Trends and Challenges” surveys key trends across parallel computing and modern distributed paradigms, and it clarifies how the two areas are related. It also summarizes open challenges and future opportunities, providing a concise roadmap for researchers and practitioners. Two additional surveys provide deep dives into important enabling layers of modern distributed systems. The survey on Transport Layer Security (TLS) 1.3 encrypted traffic analysis [1] reviews how new protocol features change what can (and cannot) be inferred from encrypted traffic and organizes state-of-the-art techniques and limitations. The software-defined wide area network (SD-WAN) survey [2] synthesizes architectures and research directions, including traffic engineering, orchestration, and security considerations in programmable wide-area networking.

Theme B—Lightweight security and authentication in distributed environments

Two papers address authentication and key agreement in settings characterized by openness, mobility, and constrained endpoints. One proposes a practical two-factor mutual authentication protocol for distributed wireless sensor networks using physical unclonable functions (PUFs), strengthening resistance to impersonation and node capture while supporting efficient session key negotiation. Another introduces a lightweight certificateless anonymous authentication and key negotiation scheme for the 5G Internet of Vehicles, targeting the security–latency requirements of high-mobility vehicular scenarios.

Theme C—Privacy-preserving distributed computation

DistOD presents a hybrid privacy-preserving distributed framework for origin–destination matrix computation, reflecting the growing need for distributed analytics tools that protect sensitive information while still supporting practical deployment and scalability.

Theme D—Learning-enabled methods for modern data processing pipelines

Two papers explore learning-driven approaches relevant to scalable decision-making and analytics workflows. One proposes a deep reinforcement learning recommendation approach based on multi-level attention mechanisms. The other presents a tensor-based method for uniform and discrete multi-view projection clustering, contributing to efficient optimization under multi-view settings.

3. Research Outlook

Across these contributions, two issues stand out. First, trustworthiness, including authentication, privacy, and operating under encrypted visibility, is increasingly central to distributed/parallel system design. Second, intelligence as an optimization tool (e.g., learning-based methods) is becoming more common across system and data pipeline contexts.

Motivated by the topics represented in this Special Issue, promising future directions include the following:

• Security and privacy by design for distributed endpoints (sensors, vehicles, Internet of Things), supported by strong threat models and practical efficiency constraints;
• Measurement and control under encryption, balancing operational needs with privacy expectations;
• Programmable wide-area infrastructures and orchestration mechanisms that enforce performance and security policies end-to-end;
• Scalable privacy-preserving analytics, especially for mobility, smart cities, and cross-organization data collaboration;
• Bridging systems and machine learning (ML), including federated and distributed learning, with methods that are reproducible, interpretable, and deployment-ready [3].

4. Conclusions

This Special Issue brings together eight papers that reflect the continued evolution of distributed and parallel computing systems to become more secure, more privacy-aware, and increasingly data- and learning-driven. We hope that this collection serves as a useful reference for the community and helps to shape future research and developments in emerging distributed/parallel computing systems.